International Journal of Trend in Scientific
Research and Development (IJTSRD)
International Open Access Journal
ISSN No: 2456 - 6470 | www.ijtsrd.com | Volume - 1 | Issue – 6
An Energy-aware
aware Distributed Clustering Protocol in
Wireless Sensor Networks
Venkateswarulu Naik. B
JJTU Scholar, Jhunjhunu, Rajasthan, India
ABSTRACT
Clustering is an effective approach for organizing a
network into a connected hierarchy, load balancing,
and prolonging the network lifetime. This paper
proposes an energy-aware aware distributed dynamic
clustering protocol (ECPF) which applies three
techniques:
1) Non-probabilistic
probabilistic Cluster Head (CH) elections.
2) Onn demand clustering. The remaining energy of the
nodes is the primary parameter for electing tentative
CHs via a non-probabilistic fashion. Anon-
probabilistic Selection is implemented by introducing
a delay inversely proportional to the residual energy
of each node. Therefore, tentative CHs are selected
based on their remaining energy. Besides, in ECPF,
CH elections are performed sporadically (in contrast
to performing it every round). Simulation resul results
demonstrate that our approach performs better than
well known protocols (LEACH, HEED, and CHEF)
in terms of extending network lifetime and saving
energy.
Keywords: Sensor Networks, Clustering, Network
Lifetime, Energy Efficient Protocols, Distributed
Algorithms
1. INTRODUCTION
Wireless sensor networks (WSNs) provide reliable
monitoring from very long distances. These networks
are basically data gathering networks in which data
are highly correlated and the end user needs a high
level description of the environ mentis ensued by the
nodes [1]. The requirements of these networks are
@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 1 | Issue – 6 | Sep - Oct 2017
Dr.S.Rama Krishna
Principal, SVIT, Secunderabad, India
ease of deployment, long system lifetime, and low-low
latency data transfers. The main task of a sensor node
in a sensor field is to detect events, perform quick
local data processing,
essing, and then to transmit the data
[2]. As mentioned in [3] and [4], nodes have typically
low mobility and are limited in capabilities, energy
supply and bandwidth. The sensor network should
perform for as long as possible. On the other hand,
battery recharging
charging may be inconvenient or impossible.
Therefore, all aspects of the sensor node, from the
hardware to the protocols, must be designed to be
extremely energy efficient [5]. In a sensor node,
energy consumption can be “useful” or “wasteful”.
Useful energy
ergy consumption can be due to one or more
of the following causes: x transmitting/receiving data
x processing query requests x forwarding queries/data
to neighboring nodes Wasteful energy consumption
can be due to:
 Idle
dle listening to the media x retransmitting
retransmi due to
packet collisions x overhearing
 Generating/handling
enerating/handling control packets [6]
In direct communication WSN, the sensor nodes
directly transmit their sensing data to the Base Station
(BS) without any coordination between the two.
However, in Cluster-based
based WSNs, the network is
divided into clusters. Each sensor node exchanges its
information only with its cluster head (CH), which
transmits the aggregated information to the BS.
Aggregation and fusion of sensor node data at the
CHs cause a significant reduction in the amount of
data sent to the BS and so results in saving both
energy and bandwidth resources. Once the clusters are
Page: 1130
 International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
constructed, each sensor node will be given an
exclusive time slot; therefore, each sensor node knows
when to transmit. Consequently, a node does not
require being awake during the complete Time
Division Multiple Access (TDMA) frame, but only
during its specific time slot [7]. To sum up, by means
of a common schedule, clustering coordinates the
transmissions of sensor nodes during the steady state
phase and so eliminates collisions, idle listening, and
overhearing. In this way, clustering achieves a
significant improvement in terms of energy
consumption. Besides, it is particularly crucial for
scaling the network to hundreds or thousands of nodes
[8]. In many applications, cluster organization is a
natural way to group spatially close sensor nodes in
order to exploit the correlation and Page | 4 eliminate
the redundancy that often shows up in the sensor
readings [9]. However, these benefits, compared to
those of the direct communication WSN, result in
extra overhead due to the cluster formation’s message
exchanges. Research on clustering in WSNs has
focused on developing centralized and distributed
protocols to compute sets of CHs and to form clusters.
Centralized approaches (e.g. [10, 11, 12, and 13]) are
rather inefficient in the case of large scale networks
since collecting the entire amount of necessary
information at the central control (BS) is both time
and energy consuming. Distributed approaches are
more efficient for large scale networks. In these
approaches, a node decides to become a CH or to join
a cluster based on the information obtained solely
from neighbors within its proximity. Several
distributed clustering protocols have been proposed in
literature (e.g. [5, 6, 14, 15, and 16]). As mentioned in
[17], most of these protocols are in either case of
iterative or probabilistic. In probabilistic protocols
(e.g. [14, 15, and 18]), the decision to become CH is
reached probabilistically. On the other hand, in
iterative protocols (e.g. [6]), the nodes perform an
iterative process to decide when there to become a CH
or not. From another point of view, clustering
protocols are considered as being static and dynamic.
In static clustering, the clusters are permanently
formed (e.g. [8, 10]), while in dynamic clustering
(like [5, 6, 9, 14, and 15]), protocol operation is
divided into rounds; clusters are formed for a round
and then should be formed again for the next round.
In doing so, extra overhead is imposed on the system.
On the other hand, some protocols (e.g. [11, 16, 12,
and 13]) take advantage of fuzzy logic. Fuzzy Logic
[19, 20] is useful for making real-time decisions
without needing complete information about the
@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 1 | Issue – 6 | Sep - Oct 2017
environment. Merging different environmental
parameters according to predefined rules and then
making a decision based on the result is another
important application of fuzzy logic. Typically, fuzzy
clustering algorithms in WSNs use fuzzy logic for
merging different clustering parameters to elect CHs.
Besides, as mentioned in [11] and [16] the overhead
of cluster head election may be highly reduced by
using fuzzy logic. This paper proposes an Energy-
aware distributed dynamic Clustering Protocol using
Fuzzy logic named ECPF. The proposed clustering
approach does not make any assumptions regarding
the distribution of the nodes or node capabilities, e.g.,
location-awareness. The protocol only assumes that
sensor nodes can vary their transmission power. In
this protocol, each node employs a process to decide
its status. For each node, this process finishes when
the node either elects itself as a CH or finds a CH to
join. Notable features of ECPF are: x Distributed CH
election (based on local information) that avoids extra
communication with the BS.
Non-probabilistic choice of CH: a node waits for a
certain delay (which is inversely proportional to the
remaining energy of that node), before it tries to
proclaim itself as a CH or join a cluster. The nodes
whose delays expire first among the neighboring
nodes will become a tentative CH, and in the next
round a cost-based choice is made to choose a final
CH from the set of neighboring tentative ones
Clustering is performed sporadically (on demand
rather than each round) when some CH depletes a
given fraction of its energy resources. We compare
our solution to LEACH, HEED, and CHEF protocols.
Simulation results in Mat lab software show that
ECPF provides superior network lifetime and energy
savings. The rest of the paper is organized as follows:
Section 2 gives a short survey of cluster based
protocols for WSNs. Section 3 describes the network
model and clustering problem. A new energy efficient
clustering scheme is outlined in Section 4. Section 5
presents the simulation results by comparing energy
consumption, network lifetime, and the number of CH
elections with other well-known algorithms. Finally,
the conclusion is presented.
2. RELATED WORKS
The following presents a review of some famous
clustering protocols. LEACH [5] minimizes energy
dissipation in sensor networks due to its constructing
of clusters. LEACH operation is performed in two
Page: 1131
 International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
phases: a setup phase and a steady state phase. In the
setup phase, a sensor node selects a random number
between 0 and 1. If this number is less than the
threshold T(n), the node becomes a CH. T(n) is
computed as:
Where r is the current round; p, the desired percentage
for becoming CH; and G is the collection of nodes not
elected as a CH in the last 1/p rounds. After being
elected, every CH announces to all of the network’s
sensor nodes that it is the new CH. When each node
receives this announcement, it chooses a cluster to
join based on the signal strength of the announcement.
The sensor nodes then inform their appropriate CH to
join them. Afterwards, the CHs, according to a
TDMA approach, assign a time slot to each node so
that its data can be sent to its CH during this period.
During the steady state phase, the sensor nodes can
perform sensing and transmit data to the CHs. The
CHs also aggregate the data received from the nodes
in their cluster before sending these data on to the BS.
After a certain period of time has elapsed in the steady
state phase, the network goes into the setup phase
again and enters the next round. The advantages of the
LEACH protocol over previous research are as
follows: In this probabilistic approach, the nodes die
randomly and at the same rate’s The dynamic
clustering of LEACH prolongs the network lifetime. x
LEACH is fully distributed and does not require
global knowledge of the network. The limitations of
the LEACH protocol are as follows: x although
energy consumption is a critical problem in WSNs,
LEACH does not consider the remaining energy of
nodes when selecting CHs. x Since CH election is
probabilistic, a node with very low energy has a good
chance of becoming a CH. When this node dies, the
entire cluster is rendered dysfunctional. x It is
possible that some CHs are located within close
proximity of each other. This indicates that CHs are
not well distributed in the network. x It is assumed
that CHs have a long communication range enabling
them to send data directly to the BS. This assumption
is not always realistic since, due to signal propagation
problems, such as the presence of obstacles, the BS is
often directly unreachable to all nodes. On the other
hand, the CHs have the capabilities of regular sensor
nodes. Consequently, LEACH is not applicable to
networks deployed in large areas.
The authors in [14] extended the LEACH’s
probabilistic CH selection algorithm. They adjusted
the threshold T(n) denoted in (1), relative to the
@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 1 | Issue – 6 | Sep - Oct 2017
node’s residual energy. Through applying this
threshold each node decides whether to become a CH
in a round or not lifetime can be efficiently increased.
There are 27 fuzzy if-then rules which are defined at
the BS. The BS elects the CHs according to these
fuzzy rules. This centralized approach is not suitable
for scalable networks because BS must collect
information about the status and location of all nodes.
LEACH-FL [2] is an improvement on LEACH
protocol which employs a similar approach to [3].
This method uses three descriptors (node residual
energy, node degree and distance from BS) for
computing the chance. The BS selects nodes with
higher chance as CHs, using 27 fuzzy if-then rules.
Although this method has the same drawback of
Gupta’s method, it presents a better result than
LEACH protocol. CHEF [4 is a fuzzy approach which
performs CH election in a distributed manner. In
every round, each node generate s a random number
between 0 and 1. If the random number is smaller
than the predefined threshold, then that node becomes
a tentative CH. There are two fuzzy descriptors that
are used in CH election: residual energy of each node
and local distance. The local distance is the sum of
distances that a node has with other nodes in radius r.
There are 9 fuzzy if -then rules that are defined in all
sensor nodes. Tentative CHs calculate their chances to
be an actual CH using these fuzzy rules. If the chance
of a tentative CH is greater than the other tentative
CHs’ chances in radius r, then that tentative CH
becomes an actual CH. Then, it sends a CH
advertisement message to the nodes in its proximity.
The nodes that are not elected as CH join the closes t
CH. This method applies a probabilistic model for CH
elections, too. Therefore, it is possible that CHs are
not well distributed in the field. Consequently, some
nodes find themselves uncovered (orphan nodes), and
have to send their sensed data directly to the BS.
Bandyopadhyay and Coyle [15] proposed another
extension of the LEACH protocol where the multi-
hop routing is applied. Similar to LEACH, every CH
advertises itself to the neighboring sensor nodes,
which relay the advertisement in a multi-hop fashion.
The advertisement is forwarded to sensor nodes in at
most h hops away. Cluster Members (CMs) that
receive multiple CH announcements, elect the closest
CH in terms of hop count. On the other hand, a sensor
node which is neither a CH nor receives any CH
announcement becomes a forced CH. WSN operation
in a multi-hop fashion has more energy conservation
in communications in comparison with single hop
Page: 1132
 International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
transmissions, especially in large scale networks. This
gain achieves in the cost of additional complexity,
e.g., the one hop CMs require data collection from the
two hop CMs. In addition, the overhead in the setup
phase increases considerably, because CH messages
have to be forwarded via multiple hops. Youngish and
Filmy [6] proposed an iterative clustering protocol,
named HEED. HEED is different from LEACH in the
manner in which CHs are elected; however, it
employs probabilistic fashion. Both electing CHs and
joining clusters are performed based on the hybrid
combination of two parameters. The primary
parameter depends on the node’s residual energy. The
alternative parameter is the intra-cluster
communication cost. This technique is utilized in
ECPF with a fuzzy cost. In HEED, each node
computes a communication cost depending on
whether variable power levels, applied for intra-
cluster communication, are permissible or not. If the
power level is fixed for all of the nodes, then the
communication cost can be proportional to (i) node
degree, if load distribution between CHs is required,
or (ii) 1/node degree, if producing dense clusters is
required. The authors defined AMRP the average of
the minimum power levels needed by all M nodes
within the cluster range to access the CH u, i.e.
MiMinPwruAMRPMi1)()(. If variable power levels
are admissible, AMRP is used as the cost function. In
this approach, every regular node elects the least
communication cost CH in order to join it. On the
other hand, the CHs send the aggregated data to the
BS in a multi-hop fashion. The advantages of the
HEED protocol are as follows:
 It is a fully distributed clustering approach that
benefits from the use of two parameters for CH
election.
 The probability of two nodes within each other’s
transmission range becoming CHs is negligible.
Therefore, in contrast with LEACH, CHs are well
distributed in the network.
 Energy consumption is not required to be uniform
for all the nodes.
 Communications in a multi-hop fashion between
CHs and the BS promote more energy
conservation and scalability in contrast with the
single-hop fashion in the LEACH protocol.
The limitations of the HEED protocol are as follows:
It uses a probabilistic model for CH elections.
Similar to LEACH, the performing of clustering in
each round imposes significant overhead on the
network. This overhead causes noticeable energy
@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 1 | Issue – 6 | Sep - Oct 2017
dissipation which results in decreasing the network
lifetime. Some protocols attempt to eliminate
overhead due to setup phase. As an example in [8],
Zhu et al presented a distributed static clustering
protocol to prolong the network lifetime. This
includes three parts. First, nodes by means of
Hausdorff clustering algorithm organize themselves
into multiple static clusters based on location of
nodes, communication effectiveness, and network
connectivity. Second, clusters are formed only once,
and the CH role is scheduled between the CMs
optimally. Third, after CH elections, CHs construct a
backbone network to periodically collect, aggregate,
and send data to the BS using minimum energy
routing. They showed that this method considerably
prolong the network lifetime in comparison with some
other known methods because it eliminates the
communication overhead due to setup phase. This
approach suffers the problems due to the proximity of
the CHs.
3. PRELIMINARIES
3.1 Network Model
The following properties are assumed in regard to the
sensor network being studied:
 The nodes can use power control to change the
amount of transmit power. Also, each node
performs signal processing functions and has the
computational power to support different MAC
protocols.
 The nodes have ideal sensing capabilities. In other
words, the quality of the node’s sensing does not
change within the cluster range regardless of the
distance from the node.
 The sensor nodes are quasi-stationary. This is
typical for sensor network applications. x Nodes
are not equipped with GPS-capable antennae,
meaning they are location-unaware In addition to
being of equal importance, the capabilities of
nodes, such as processing and communicating, are
similar.
 Nodes are energy constrained and are left
unattended after deployment. Therefore, battery
recharge is not possible.
 Because the energy consumed per bit for sensing,
processing, and communicating is typically
known, remaining energy can be estimated. As a
result, measuring this remaining energy is not
essential.
Page: 1133
 International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
 Each node has an initial amount of energy, Amax,
and the BS is not limited in terms of energy,
memory, and computational power.
 Node failures are basically due to energy
depletion.
 Distance can be measured based on the wireless
radio signal power.
 Links are symmetric, i.e., two nodes v1 and v2
can communicate using the same transmission
power. TCP and TNO are defined as follows:
 TCP (the period of the clustering process) is the
time interval used by the clustering protocol to
cluster the network.
 TNO (the network operation interval) is the time
between the end of a TCP interval and the start of
the subsequent TCP interval. In order to reduce
overhead, it must be ensured that T NO TCP.
Note that, in contrast with other dynamic clustering
protocols that perform clustering in each round, ECPF
clusters the nodes on demand rather than at each
round. Therefore, it is possible that some rounds do
not include TCP; instead TNO is extended during
these rounds. As a result, the length of the TCP
interval is fixed but the length of the TNO interval
varies throughout the network lifetime. In ECPF, node
readings are periodically reported to the BS.
Therefore, a TDMA frame is created in each CH to
remove interference within a cluster. The protocol
uses special synchronization pulses to alert the sensor
nodes that clustering will be triggered in the
beginning of the next round. These pulses are
propagated in a centralized multi-hop fashion (like the
approach presented in [21]). The basis of this
approach is the construction of a low-depth spanning
tree T comprising the nodes in the network. In
general, a new spanning tree is constructed each time
the algorithm is performed. In order to synchronize
nodes in the tree, pair-wise synchronizations are
performed along the edges of T. In centralized multi-
hop synchronization, the reference node (BS) initiates
the synchronization through all its immediate (single-
hop) children in T. Next, every child of the reference
node, in turn, synchronizes with its children. This
process continues until the leaf nodes of T are
reached. The algorithm terminates when all the leaf
nodes are synchronized. The running time of the
algorithm is proportional to the depth of the tree
which is log(n), where n is the number of nodes in the
tree. Therefore, these pulses are quickly penetrated
throughout the network. 3.2 The Clustering Problem
Suppose the above assumptions hold and that n nodes
@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 1 | Issue – 6 | Sep - Oct 2017
are distributed in a field. In consideration of energy
saving issues, the goal is to identify a collection of
CHs which cover the entire area. Each node vi, where
1 ≤ i ≤ n, must be mapped to exactly one cluster cj,
where 1 ≤ j ≤ nc, and nc is the number of clusters (nc
≤ n). Li shall denote the lifetime of node i. The
network lifetime will be defined as follows: x F is the
time elapsed until the First Node Dies (FND).
Therefore, F = min (L1, L2, …, Ln). x H is the time
elapsed until only one Half of Nodes remain Alive
(HNA). In other words, H = median (L1, L2, …, Ln).
x L is the time elapsed until the Last Node Dies
(LND) or, L = max (L1, L2, …, Ln). The major
purpose here is to maximize F, H, and L, which
requires using the energy of all nodes uniformly. A
node must have the ability to directly communicate
with its CH and by a single-hop fashion. A CH has
two critical responsibilities: (1) intra-cluster
coordination and (2) inter-cluster communication.
Multi-hop routing is used for inter-cluster
communication. CHs can utilize a routing protocol to
compute inter-cluster paths for communicating in a
multi-hop fashion with the BS, e.g. the power-aware
routing protocol in [22].
The following requirements are recommended:
 Clustering is fully distributed. Each node decides
independently based on local information’s
Clustering finishes within a fixed number of
iterations (regardless of network diameter).
 At the end of each TCP, each node is either a CH
or a regular node that belongs to exactly one
cluster.
 In terms of processing complexity and message
exchange, clustering should be efficiently
performed.
 CHs are well distributed over the sensor field.
Note that, in the clustering process, every iteration
takes time, tc. Period tc should be long enough to
receive messages from any neighbor within the
cluster radius. Because the nodes are quasi-
stationary, neighbor discovery is not required
every time clustering is performed. Therefore, the
neighbor set of every node does not vary very
frequently. In multi-hop networks, the nodes
automatically update their neighbor sets by
periodically sending and receiving heartbeat
messages.
3. THE PROTOCOL
In this section, the ECPF and its pseudo code are
illustrated. The operation of ECPF is divided into
rounds and each round is comprised of two phases:
Page: 1134
 International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
1. The setup phase, which includes CH election and
consequently cluster formation. In addition, in this
phase, every CH coordinates with its members to
send sensing data during the following phase.
2. The steady state phase, which is broken up into
TDMA frames. During each frame, every regular
node, at the time of its respective time slot, sends
sensing data to its CH (similar to [5]). At the end
of each TDMA frame, every CH forwards the
aggregated data to the BS through the CHs.
This protocol has the following characteristics which
resemble [6]’s:
 The steady state phase is similar.
 A chosen CH advertises only to its neighbors.
 Each node can directly communicate with its CH.
 During the clustering process, a node can be either
a tentative_CH or a final_CH or it can be covered
At the end of the setup phase, CHs form a network
backbone, such that packets are routed from the
CHs to the BS in a multi-hop fashion over CHs.
4.1 On Demand Clustering
A novelty of ECPF is that it decreases overhead by
performing the setup phase on demand instead of in
each round, To do so, when the clustering process
finishes (at the end of each setup phase), every CH
saves its residual energy in its memory, for example
in its ECH variable. During the steady state phase,
whenever a CH finds that its Residual falls below
echo (α is a constant number and 10, it sets a
prespecified bit in a data packet which is ready to be
sent to the BS in the current TDMA frame. Upon
receiving the forwarded CH data packet (sent in a
multi-hop fashion), the BS informs the sensors to hold
the setup phase at the beginning of the upcoming
round. This could be achieved by having the BS send
out, in a multi-hop fashion, specific synchronization
pulses to nodes. These pulses are quickly dispersed
throughout the network. When every node receives a
pulse, it prepares itself to perform clustering.
Therefore, CH election and consequently cluster
formation are performed on demand. As a result, the
overhead created by consecutive setup phases is
tremendously reduced. Consequently, there is a
decrease in the energy dissipation of nodes and an
increase in network lifetime.
4.2 Fuzzy Cost
Fuzzy Logic (FL) is used to model human experience
and human decision making behavior. In FL the
@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 1 | Issue – 6 | Sep - Oct 2017
input-output relationship is expressed by using a set of
linguistic rules or relational expressions. As shown in
Fig. 2 a FL basically con sits of four important parts
including a fusilier, a defuzzifier, an inference engine,
and a rule base. As in many fuzzy applications, the
input data are usually crisp, so a fuzzification is
necessary to convert the crisp input data into a
suitable set of linguistic value which is needed by the
inference engine. In the rule base of an FL, a set of
fuzzy rules, which characterize the dynamic behavior
of the system, are defined. The inference engine is
used to form inferences and draw conclusions from
the fuzzy rules. The output of the inference engine is
sent to the defuzzification unit. Defuzzification is a
mapping from a space of fuzzy actions into a space of
crisp actions. We have employed the most commonly
used fuzzy inference technique, called the Madman
[23] method, because of its simplicity. To obtain a
cost, ECPF uses two fuzzy sets and the fuzzy if-then
rules. We adjusted the input variables, used in the
fuzzy if-then rules, between 0 and 1 such that the
fuzzy sets will be applicable for any size of networks.
The fuzzy system input variables are defined as
follows:
 Node degree: the number of neighbors a node has
which is divided by total number of nodes in the
network. In other words, node_degreei =
(|Snarl(i)| / #nodes) where, Snarl(i) = {v: v lies
within node if’s cluster range}.
 Node centrality: a value that shows how central
the node is among its neighbors proportional to
network dimension, or
node_centralityi=Dimension
NetworkiSjidistnbriSjnbr_)(/)),(()(2¦.
Since, typically transmissions energy is proportional
to the squared distance, the lower value of the node
centrality results in a lower amount of energy required
by other nodes to send data to the node assuming the
role of a Chute fuzzy sets of input variables and
output variable (cost) are described in Fig. 3. ECPF
calculates a cost using fuzzy if-then rules. A smaller
cost means that the node has a higher priority of being
elected a CH. Based on the two fuzzy variables, fuzzy
if-then rules can be defined which are similar to those
presented in Table I. After aggregating the results
achieved from each rule, a efuzzification method is
required to obtain the crisp value. Defuzzification is
performed using the Coal method, which returns the
Center of Area under the fuzzy set achieved
aggregating conclusions. demonstrates the effect of
nodes’ attributes on the chance of the nodes becoming
Page: 1135
 International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
CHs. This plot was generated by the rules that
accounted for both node degree and node centrality
factors. With the increase of node degree and decrease
of node centrality, the cost of a node to be selected as
a CH is decreased. A node with lower cost has more
priority for becoming CH.
4.3 Clustering Process
The clustering process of ECPF is divided into three
phases:
Initialization Phase:
In the beginning of this phase, neighbor information
could be updated using CSMA/CA. Afterwards, each
node may compute its cost independently. The cost is
the output of fuzzy system described in the previous
section. This cost will not be broadcasted to neighbors
as it is exchangeable through Chums messages. As
previously mentioned, note that updating the neighbor
information and computing costs are not required
every time clustering is triggered. Each sensor node
sets its own delay time:
),(1_maxmaxdEEMAXtimedelayresidualmIn the
above, Residual is the current energy of the sensor
node and Exam is the maximum energy
corresponding to a fully charged battery. Therefore,
the main constraint in sensor nodes, i.e., residual
energy, is taken into account in this delay time. The
value of dams limits the period of time should be
elapsed in the clustering process. See the Initialization
phase in x Main processing Phase: In this phase, every
node must wait until the expiration of its delay time.
If a node does not hear the Chums message from any
other sensor node during the delay time, upon delay
time expiration it shall declare itself to be a tentative
CH. The node announces its status by sending
Chums(Nodded, tentative_CH, cost), to all the nodes
within the cluster range. Note that when a node has
higher energy, its delay time is less than that of nodes
with a lower amount of energy. As a result, because
its delay time has expired sooner, the node has a
higher priority of becoming selected as a
tentative_CH. In this way, the non-probabilistic
method of selecting CHs is employed. In the next
iteration, if this particular node has the least cost
among the tentative_CHs in its proximity, it will
become a final_CH and shall broadcast a final_CH
message within its cluster range. On the other hand, if
a node receives a final_CH message, it can no longer
be elected as a CH. Therefore, in the following phase,
it must choose to connect to one of the final_CHs in
its cluster radius, based on the cost of that final_CH.
Observe that, in this phase, each tentative_CH (or
@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 1 | Issue – 6 | Sep - Oct 2017
final_CH) node can send a Chums only once. It is
possible that in the beginning of this phase some
neighboring nodes whose delays expire at the same
time will become tentative_CHs, however in the next
iteration only one of them declares itself as a
final_CH. Therefore, final_CHs do not locate in each
other’s cluster range. See the Main processing phase x
Finalization Phase: During this phase, each sensor
node makes a final decision about its status. If the
node is not a final_CH and has received at least one
final_CH message, it will elect the final_CH with the
least cost to join it. If a node completes the clustering
process and has not yet received any final_CH
message, it will find itself uncovered and so shall
introduce itself as the final_CH. See the Finalization
phase .All distributed protocols face the convergence
issue in the cluster head election (e.g. when two nodes
receive tentative CHs messages with same cost). This
condition rarely happens; both of the nodes should
have equal amount of energy to have equal delay
time, on the other hand, they should not be in the
cluster range of any final_CH. Besides, they should
execute the algorithm at the same time. However, the
least cost function can select the node with the lower
ID in this case.
5. SIMULATION RESULTS
In this section, a comparison between the simulation
results in ECPF, LEACH, HEED and CHEF protocols
is performed via Mat lab software. The following
assumptions and system parameters (similar to [5])
are used:
 The nodes always have data to send to the end
user and the nodes situated in close proximity to
others have correlated data.
 The energy required for data aggregation is set as
EDA=5nJ/bit/signal and CHs perform ideal data
aggregation (i.e. all the messages received from
cluster members can be aggregated into a single
message).
 A simple model for the energy dissipation of radio
hardware is assumed, in which the receiver
dissipates energy to run the radio electronics and
the transmitter dissipates energy to run the power
amplifier and radio electronics, as shown in. Thus,
for transmitting a k-bit message over distance d,
the radio expends: ETx(kid) = ETx_elec(k) +
ETx_amp(kid)
Page: 1136
 International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
5.1. Setting α Variable
As mentioned before, CH election (clustering
execution) is performed on demand. When a CH
consumes a prespecified part of its energy (i.e.
CHresidualEEDd), it indirectly informs the other
nodes that clustering must be performed for the
upcoming round. In order to obtain the proper α,
ECPF was run for the two scenarios described above.
α differs from 0 to 1 and each plot demonstrates the
average of three executions for 100, 200, 300, and 400
number of nodes, until the first node dies (FND). Also
plotted is an average of the four mentioned plots.
When α is equal to zero, no clustering is performed
during the network lifetime (i.e. the static clustering
approach that considers fixed CHs). In homogenous
networks in which nodes have similar capabilities and
the same amount of energy, CHs quickly deplete
respective energies. Therefore, when a CH dies, the
respective cluster becomes dysfunctional. When α
equals one, this signifies that clustering is performed
in each round, similar to the LEACH, HEED and
CHEF protocols. Considering the plot, which is the
average of the four different numbers of nodes,
figures 8-14 are plotted using 8.0 Das it
approximately results in better network lifetime.
5.2 Energy Consumption Comparisons
In this subsection, the energy dissipation to cluster the
WSN and the energy consumption to transmit the
sensed data to the BS are evaluated. Note that in the
figures belonging to this subsection, the vertical axis
in Scenario 1 does not have an equal range of data in
contrast to Scenario 2 because of the high energy
dissipation of LEACH in Scenario 2. In Fig. 8, the
average energy dissipation of protocols clustering the
WSN per election is evaluated. The ECPF performs
better because its clustering’s message complexity is
low and, similar to HEED and CHEF.
6. CONCLUSION
In this paper, we proposed an energy efficient,
distributed clustering protocol for WSNs. Our
approach can be useful for applications that require
scalability, prolonged network lifetime and nodes are
dispersed in a spacious field. We assumed quasi-
stationary networks in which nodes are location-
unaware and have equal significance. Based on this
assumption, we presented the ECPF, where terminates
CH election process with a constant number of
iterations, and without any dependency of the network
@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 1 | Issue – 6 | Sep - Oct 2017
diameter. Combining fuzzy logic, on demand
clustering, non-probabilistic CH election, and
consideration of nodes’ energy, allows ECPF achieve
longer lifetime when compared to existing clustering
protocols. Many applications require the ability to
provide information from each part of the monitored
area at any moment in order to meet the application’s
quality of service (Quos) [9]. For future works, we
would like to extend the protocol to meet Quos
requirements of WSNs, such as coverage
preservation, because complete coverage of the
monitored area over long period of time is an
outstanding issue.
REFERENCES
1) W. B. Heinemann, “Application-specific protocol
architectures for wireless networks,” Ph.D. Thesis,
Massachusetts Institute of Technology, (2000).
2) I. Akiyldiz, W. Su, Y. Sankarasubramaniam, and
E. Cerci, “A survey on sensor networks,” IEEE
Communication Magazine, vol. 40, no. 8, pp. 102-
114, (2002).
3) O. Youngish and S. Foamy, “Distributed
clustering in ad-hoc sensor networks: A hybrid,
energy-efficient approach,” in Proceeding of IEEE
INFOCOM, vol. 1, pp. 629-640, (2004).
4) K. Akaka, M. Youngish, “A survey on routing
protocols for wireless sensor networks,” Ad Hoc
Networks, vol. 3, no. 3, pp. 325-349, (2005).
5) W. B. Heinemann, A. P. Chandrakasan, and H.
Balakrishnan, “An application-specific protocol
architecture for wireless micro sensor networks,”
IEEE Transactions on Wireless Communications,
vol. 1, no. 4, pp. 660-670, (2002).
6) O. Youngish, S. Filmy, “HEED: A Hybrid,
Energy-Efficient, Distributed clustering approach
for Ad Hoc sensor Networks,” IEEE Transactions
on Mobile Computing, vol. 3, no. 4, pp. 366-379,
(2004).
7) N. Bouabdallah, M. E. Rivers–Angeles, and B.
Seri cola, “Continuous Monitoring Using Event-
Driven Reporting for Cluster-Based Wireless
Sensor Networks,” IEEE Transactions on
Vehicular Technology., vol. 58, no. 7, pp. 3460-
3479, (2009).
8) X. Zhu, L. Sheen, and T. P. Yum, “Hausdorff
Clustering and Minimum Energy Routing for
Wireless Sensor Networks,” IEEE Transactions on
Page: 1137
 International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
Vehicular Technology, vol. 58, 2, pp. 990-997, 19) L. A. Zane, “Fuzzy Sets,” Information and
(2009). Control, vol. 8, 3, pp. 338-353, (1965).
9) S. Sorrow, W. Heinemann, “Cluster head election 20) L. A. Sade, “Outline of a New Approach to the
techniques for coverage preservation in wireless Analysis of Complex Systems and Decision
sensor networks”, Ad Hoc Networks, vol. 7, no. 5, Processes,” IEEE Transactions on Systems, Man
pp. 955-972, (2009). and Cybernetics, SMC-3, 1, pp. 28-44, (1973).
10) T. Kauri, J. Beak,” A Strategic Deployment and 21) J. Greened, J. Rabies, “Lightweight Time
Cluster-Header Selection for Wireless Sensor Synchronization for Sensor Networks,”
Networks,” IEEE Transactions on Consumer International Workshop on Wireless Sensor
Electronics, Vol. 55, 4, pp. 1890-1897, (2009). Networks and Applications, (2003).
11) I. Gupta, D. Riordan, and S. Sample, “Cluster- 22) S. Singh, Miwok, and C. Raghavendra, “Power-
head election using fuzzy logic for Wireless aware routing in mobile ad hoc networks,” in
Sensor Networks,” In Communication Networks Proceeding of 4th Annual ACM/IEEE
and Services Research Conference, Proceedings of International Conference on Mobile Computing
the 3rd Annual, pp. 255–260, 2005. Networking (Modicum), (1998).
12) G. Ran, H. Zhang, S. Gong, “Improving on 23) E.H. Madman, and S. Assailant, An experiment in
LEACH protocol of Wireless Sensor Networks linguistic synthesis with a fuzzy logic controller.
Using Fuzzy Logic,” Journal of Information & International Journal of Man-Machine Studies,
Computational Science, pp. 767-775, 2010. vol. 7, 1, pp. 1-13, (1975).
13) N. Bazaars Torghabeh, M. R. Akbarzadeh Toto 24) R.Shah, J. Rafael, “Energy aware routing for low
chi, and M. H. Yaghmaee Moghaddam, “Cluster energy ad hoc Sensor Networks,” IEEE Wireless
Head Selection using a Two-Level Fuzzy Logic in Communications and Networking Conference, pp.
Wireless Sensor Networks,” International 350-355, vol.1, (2002).
Conference on Computer Engineering and
Technology, pp. 357-361, 2010.
14) M. C. M. Thin, T. Their, “An Energy Efficient
Cluster-Head Selection for Wireless Sensor
Networks,” International
15) S. Bandyopadhyay and E. J. Coyle, “An energy
efficient hierarchical clustering algorithm for
wireless sensor networks,” in Proceeding of IEEE
INFOCOM, vol. 3, pp. 1713-1723, (2003).
16) J. M. Kim, S.H. Park, Y.J. Han, and T.M. Chung,
“CHEF: cluster head election mechanism using
fuzzy logic in Wireless Sensor Networks,”
International Conference of Advanced
Communication Technology, pp. 654–659, 2008.
17) O. Youngish, M. Runs, and S. Ramasubramanian,
“Node Clustering in Wireless Sensor Networks:
Recent Developments and Deployment
Challenges,” IEEE Network, vol. 20, 3, pp. 20-25,
(2006).
18) W. B. Heinemann, A. P. Chandrakasan, and H.
Balakrishnan, “Energy-Efficient Communication
Protocol for Wireless Micro sensor Networks,”
Proceeding of the Hawaii International
Conference on System Sciences, pp. 1-10, (2000).

@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 1 | Issue – 6 | Sep - Oct 2017 Page: 1138